- Boost - lists.boost.org

forwarding-function macro library?
by Isaac Dupree 12 Aug '08

12 Aug '08

Is there a macro library in Boost for making forwarding functions (the thing that can only be done perfectly in C++0x, with decltype, rvalue references and perhaps parameter packs)? I found myself needing to do that again, as I worked more on my ideas for integer wrapping, and so I started making a macro library for it... I have no idea where I'd look to find if such a thing already existed, beyond reading the summaries of all the Boost libraries. I suppose it would be helpful to have a general library in Boost to do that? (with all the compiler workarounds, the choice of the user which workaround to prefer, etc.) (more generally/irrelevantly, I'm surprised how often Boost libraries share techniques that ought to be libraries in themselves. And how poorly commented some of Boost code is... but there I'm getting way off topic :-P) -Isaac

3 3

[1.36.0 beta][serialization] One more typo
by Ivan A. Melnikov 12 Aug '08

12 Aug '08

In 1.36.0 beta, in file boost/serialization/detail/shared_ptr_132.hpp, an ifdefed-out line 20 states: #include <boost/serializaition/detail/shared_ptr_nmt_132.hpp> Note misspelled word: serializaition. Maybe we could add some simple test (like those that check for propper line endings) to check for such typos? At least for includes that start with boost/... -- Best Regards, Ivan A. Melnikov <iv(a)altlinux.org>

1 0

[preprocessor] Reverse file iteration
by Robert Dailey 12 Aug '08

12 Aug '08

Hi, Is it possible to do a reverse iteration? For example: # define BOOST_PP_ITERATION_LIMITS (5, 0) Looking at the preprocessor output from MSVC9, I'm seeing that it works properly. By properly, I mean I'm getting my "primary" template first, and then several specialized versions under that one. I'm using reverse iteration here so that I can create class specializations, while at the same time keeping my primary template at the "top", since it will be the first to be created. However, any code that tries to use a specialized version of my class does not compile. It says insufficient number of template parameters, which leads me to believe iteration limits does not expect start to be greater than finish. Any advice here? I'll post code if needed.

2 1

[units][1.36.0 beta] A bug?
by Ivan A. Melnikov 12 Aug '08

12 Aug '08

In boost 1.36 beta, in file boost/units/homogeneous_system.hpp line 21 states: #include <boost/is_same.hpp> while no such file seem to exist. Is it a bug or am I missing something? -- Best Regards, Ivan A. Melnikov <iv at altlinux dot org>

5 5

Updating news feeds
by Martin Vuille 11 Aug '08

11 Aug '08

Hi, I am trying to update the news feed on the website by following the instructions at http://www.boost.org/development/website_updating.html In step two, bjam reports that the python script failed, but there is no error message from the script itself. In fact, if I run the script manually with the same arguments as bjam, it works fine and returns a zero exit status. Am I missing something obvious? MV

1 0

[1.36] Ready to ship
by Beman Dawes 11 Aug '08

11 Aug '08

The plan to start packaging 1.36 for release tomorrow (Saturday). The beta went smoothly enough that there seems little benefit of further betas or release candidates. Please do not may any further branches/release code commits without checking with me first. Please do not make any documentation commits after 11::00 AM UTC Saturday (6 AM US Eastern Daylight Time). Thanks, --Beman

2 2

[ANN] Boost.Proto committed to trunk
by Eric Niebler 11 Aug '08

11 Aug '08

I've just committed Boost.Proto v4 to trunk (http://svn.boost.org/trac/boost/changeset/48079) This version is represents a significant reworking from the version that was reviewed and accepted back in April. Although the code is quite stable now, I have not yet integrated all the feedback about the documentation. It is my intention to bring the docs up to Boost standards in time to get Proto into 1.37. I am leaving the old version of Proto at boost/xpressive/proto for the time being, since some other projects (Xpressive, Spirit2 Qi and Karma) are relying on it. It will be removed at the earliest opportunity. I've also updated the proto.zip archive in the Boost file vault and uploaded the latest docs to: http://boost-sandbox.sourceforge.net/libs/proto/doc/html/index.html Thanks, -- Eric Niebler BoostPro Computing http://www.boostpro.com

1 0

proposal: interval containers; boost examples and more
by Joachim Faulhaber 11 Aug '08

11 Aug '08

Dear boost developers, after encouraging feedback by Paul A Bristow, Vicente Botet and Luke Simonson to my proposal on interval containers from end of May (http://lists.boost.org/Archives/boost/2008/05/137301.php) I started to boostify my interval template library (ITL). Even though there is a lot left to be done for the ITL to become fully boost compliant I'd like to report on some aspects of my current work. (1) As suggested by Paul A Bristow I have prepared some new examples that are using boost::date_time as instance types for the interval containers. * Example boost_party.cpp is a boostified instance of the introductory example party.cpp that demonstrates the 'aggregate on overlap' mechanism of interval maps (see: http://www.herold-faulhaber.de/itl/boost__party_8cpp.html) * Example month_and_weeks_grid.cpp demonstrates how different time grids can be merged in split_interval_sets by intersection (see: http://www.herold-faulhaber.de/itl/month__and__week__grid_8cpp.html) * Example man_power.cpp shows how calculations on interval_sets and maps can be obtained using set type operations. In the example the available man power in a small company is calculated accounting for weekends, holidays vacations etc. (see: http://www.herold-faulhaber.de/itl/man__power_8cpp.html) * Example user_groups shows how interval maps can be unified or intersected (see: http://www.herold-faulhaber.de/itl/user__groups_8cpp.html) (2) In the course of boostification of my code I have reviewed the design and completed the interval container classes as follows: * interval_set * split_interval_set * separate_interval_set (added) * interval_map (added) * split_interval_map You can see an overview these interval containers and some basic characteristics in the ITL's documentation here: http://www.herold-faulhaber.de/itl/interval__container__conduct_8h.html The major addition here is the class template interval_map. In contrast to a split_interval_map that preserves all interval borders that may emerge due to insertion and intersection operations, an interval_map keeps itself always minimal: Whenever two neighbouring interval-value pairs ([a,b), v), ([b,c),v) emerge, they will be joined to ([a,c),v). (3) Reviewing the overall design of the interval containers I have separated the meaning of insert/add and erase/subtract, which had been confounded in the previous version of the ITL. This has consequences especially for interval maps: As a design decision I'd like to choose that interval maps are addable and subtractable objects having operators += and -= that implement these properties. Addition and subtraction is then defined via element addition (add(value_type&)) and element subtraction (subtract(value_type&)). At a general level I can define add<inplace_op>(value_type&) and subtract<inplace_op>(value_type&)as member function templates, that implement aggregation on overlap (aggovering) not only for inplace_plus (+=) but also for arbitrary aggregate operations (inplace_op). It turned out, that function insert (std::insert semantics) can be expressed by the general add template (as well as erase can expressed by the subtract member function template): add<inplace_identity>(value_type&) is equivalent to insert(value_type&) subtract<inplace_erasure>(value_type&) is equivalent to erase(value_type&) In addition we can also define 'plain' element add and subtract by means of these templates: add(value_type&) can be defined as: add<inplace_plus>(value_type&) subtract(value_type&) can be defined as: subtract<inplace_minus>(value_type&) So for the ITL container classes I do provide now insert and erase that covers the std::semantics and I provide add and subtract that is the basis for addability and subtractability of itl::maps and itl::interval_maps. You can find an overview over all insert/erase and add/subtact functions and their relationships to std::semantics in the ITL's documentation here: http://www.herold-faulhaber.de/itl/interval__container__design_8h.html I find these design and semantical questions quite rich, and sometimes surprising ;-) and I have found many more aspects and properties while reviewing the ITL's design. I am planning to prepare a more detailed article about these topics in the next months to come. (4) I have prepared a new release 2.0.1 of the ITL for the current state that can be downloaded from: http://www.sourceforge.net/projects/itl Documentation is availabe online also from here: http://www.herold-faulhaber.de/itl The code is well tested using the Law Based Test Automaton (LaBatea) and is protabel for gcc 3.4.4 and msvc 8.0 sp1. I am sorry the ITL is still not in the boost standard directory structure. This will be one of the next steps on my todo list. (5) Concerning boostification of ITL there is still a lot to be done. Besides reviewing of the overall design reported above, I have removed virtual functions from all interval container classes using the curiously recurring template pattern (CRTP). To prevent you repeating advices that already found their way into my todo list I am listing it roughly as appendix to this posting. Cheers Joachim Faulhaber --- Boostification todo list: - Extracting a subset of the ITL as boost proposal - Providing the code in boost directory structure - Replacing own metaprogramming code by boost::mpl,fusion etc. - Using boost::value_initialized for generic initialisations - Further minimizing of class template interfaces and... - Providing more operators and functions as itl::global algorithms - Providing as much std::semantics as possible for interval containers (viewed as set<interval<T>> or map<interval<T>,U>) - Adapting testcode to boost testing standards - Improving documentation especially for issues of design and complexity - Providing boost style documentation - Preparing a proposal on the addition of open interval bounds to boost::interval

4 6

[GIL] read/write PNG/JPEG to/from istream/ostream
by Daniel Pfeifer 11 Aug '08

11 Aug '08

Hello Boosters, Find a modified version of the IO extension of Boost.GIL uploaded to the vault where I added functionality to read from std::istream& and write to std::ostream& for both PNG and JPEG files. Get it from http://tinyurl.com/gil-iostream Cheers, Daniel

3 2

Crypto in Vault...
by Dean Michael Berris 11 Aug '08

11 Aug '08

Hi Guys, Is anybody actively developing the Crypto library implementation in the vault? I ask because I'd like to know if there would be a possibility to make it a header-only implementation. This should make it a better/easier fit for adding cryptographic hash support to Boost. Have a good day everyone. -- Dean Michael C. Berris Software Engineer, Friendster, Inc.

2 4